Active Galactic Nuclei

Outline
●
Early history
●
AGN basics
●
Finding AGNs
●
AGN terminology
●
AGN unification
●
Dissecting AGNs
●
Focused lecture: Proving the Evidence of a
SuperMassive BH
Some references
 Some references

●
Plus tons of online resouces:
– Monaco
– Marconi
– Risaliti
– Brandt
– Xie
– Etc..
 Outline
●
Early history
●
AGN basics
●
Finding AGNs
●
AGN terminology
●
AGN unification
●
Dissecting AGNs
●
Focused lecture: Proving the Evidence of a
SuperMassive BH
Early History of AGNs
 Early History
●
1908 – Edward Fath notices strong emission lines from H, O, Ne in the nuclear
spectrum of NGC 1068. This was way before:
– 1915 – General Relativity
– 1916 – Schwarzschild solution found, but not fully understood
– Early 20s realizing that galaxies were extragalactic objects!!
●
1917 – Vesto Slipher obtains a higher quality spectrum of NGC 1068 and notes its
emission lines are unusually broad
●
1918 – Herber Curtis notes in M87 a “curious straight ray ... connected with the nucleus
by a thin line of matter”
●
1924-1929 – General realization that galaxies are extragalactic – led by Edwin Hubble
●
1926 – Edwin Hubble notices the nuclear emission-line spectra of NGC 1068, NGC
4051, NGC 4151
 Early History
●
1939 – Grote Reber discovers the radio source Cygnus A
●
1943 – Carl Seyfert shows that a fraction of galaxies have strong, broad emission lines and
that these galaxies are especially luminous – now known as “Seyfert galaxies”
●
1954 – Walter Baade and Rudolph Minkowski find the counterpart to Cygnus A at z = 0.057
●
1963 – Maarten Schmidt discovers 3C273 to have z = 0.158
●
1964 – Zeldovich & Novikov and Salpeter speculate about black holes powering quasars
●
1967 – The term “black hole” comes into general use
●
1968 – Donald Lynden Bell notes that many galactic nuclei may contain “collapsed old
quasars”
●
After – AGNs become a topic of widespread study
 Outline
●
Early history
●
AGN basics
●
Finding AGNs
●
AGN terminology
●
AGN unification
●
Dissecting AGNs
●
Focused lecture: Proving the Evidence of a
SuperMassive BH
 AGN basics:
What are Active Galaxies?
●
In the local
universe…
– 10-6 of massive
galaxies contain
luminous quasars
– ~ 5% are moderately
luminous (Seyfert
galaxies)
– ~ 30% show signs of
low-level nuclear
activity
 Seyfert Galaxies
●
Carl Seyfert (40s)
●
Peculiar spiral galaxies with strong emission lines in the
nuclei

●
Emission lines indicates an ionization
level of the gas larger than HII regions
and
●
Emission lines can have FWHM>1000
km/s (normal galaxies ~ few 100s km/s)
 Ultraviolet Spectrum of
Seyfert Galaxies

Carbon
Hydrogen

Oxygen

NGC 4151:
Digital Sky Survey

UV spectrum of the Seyfert galaxy NGC 4151

observed with Hopkins UV Telescope
 NGC 1068:
An Obscured Seyfert

This bright cloud has a direct

view of the intense nucleus

Radiation and (possibly) a wind

escapes through a cone-shaped
funnel

Nucleus is hidden in here

 Basic Observed
Properties of AGNs
Five main observed properties

●
Broad range of luminosities, reaching very large values.
●
Strong and broad optical/UV emission lines.
●
Emission over a very broad band (very unlinkely from
stars).
●
Variability.
●
Particle jets.
 Broad Range of Luminosities

Maoz (2007)

Steffen et al. (2006)

Low-luminosities

Optical vs
X-ray Power Intermediate
(Seyfert)
Quasars
Broad Range of Luminosities
 Broad Range of Luminosities
●
Span 9+ orders of magnitude in luminosity.
●
There is no strict lower limit on luminosity; e.g., even the black hole
at the center of our Galaxy shows some intermittent activity at very
low levels (comparable with 1033 in the previous units.
●
At very low luminosities, the distinction between active and normal
galaxies is largely semantic. There is no clear bimodal separation of
properties, for example.
●
There is a maximum observed luminosity, and we believe that we
have found examples of the most luminous AGNs that exist (i.e., the
most luminous quasars).
●
They outshine their own galaxy by factors of 100-1000s.. Can easily
be detected to high-z
 Strong and broad optical/UV
emission lines.
Vanden Berk et al. (2001)

●
Stacked Spectra from SDSS (“average
spectra”)
●
Dashed approx continuum
●
Strong and broad optical/UV emission lines
●
Almost like a forest of ELs
 Strong and broad optical/UV
emission lines.
What is telling us?
●
Indicates the presence of ionized nebular gas.
●
Gas is photo-ionized by the bright central
source.
●
Line widths indicate high-speed motions with a
wide range of velocities, up to ~ 25000 km s -1 !!
Not thermal motion, but
doppler → bulk motions
 Strong and broad optical/UV
emission lines.
What is telling us?
●
Indicates the presence of ionized nebular gas.
●
Gas is photo-ionized by the bright central
source.
●
Line widths indicate high-speed motions with a
wide range of velocities, up to ~ 25000 km s -1 !!
Some lines are narrower [O III] vs Hβ
●
Broad-line and Narrow-line
regions of AGNs
 Strong and broad optical/UV
emission lines.
What is it telling us?

●
Indicates the presence of ionized nebular gas.
●
Gas is photo-ionized by the bright central source.
●
Line widths indicate high-speed motions with a wide
range of velocities, up to ~ 25000 km s -1 !!
●
Abundances about solar or slightly super-solar (very
difficult measurements).
●
(Sometimes) AGNs present blue-shifted absorption lines
 Blueshifted absorption

What is it telling us?

●
Indicate the presence of ionized nebular gas.
●
Gas is photo-ionized by the bright central source.
●
Line widths indicate high-speed motions with a wide
range of velocities, up to ~ 25000 km s -1 !!
●
Abundances about solar or slightly super-solar (very
difficult measurements).
●
(Sometimes) AGNs present blue-shifted absorption lines

●
Can be very broad (large velocities-thousands km/s)
●
Indicate the presence of absorbing gas along the LOS
Emission over a very broad band

~10%

~90%

Typically much flatter spectrum w.r.t. e.g. stellar emission

 Emission over a very broad band

~10%

Ho et al. 2008
Emission over a very broad band

~10%

~90%

Typically much flatter spectrum w.r.t. e.g. stellar emission

 Variability of AGNs

What is it telling us?

●
Indicates the presence of ionized nebular gas.
●
Gas is photo-ionized by the bright central source.
●

100 years of 3C 273

The variability is generally chaotic without a clear period or quasi-period.
 Variability of AGNs
H band

● V band
●

2 keV

●
Variability of 3C 273 at Many Wavelengths
●
The variability is generally chaotic
●
No a clear period or quasi-period
●
Variability in different bands is often correlated.
 Variability of AGNs

●
Variability increases (larger amplitude and more rapid) as one moves to higher
frequencies.
●
Rapid X-ray Variability on Timescales Down to Minutes!!
●
X-ray variability often implies an emission-region size of light hours or less.
 Variability of AGNs

●
Rapid X-ray Variability on Timescales Down to Minutes!!
●
X-ray variability often implies an emission-region size of light hours or less.
 Variability of AGNs

Long-Term 2-10
keV Monitoring
of Five Seyfert
Galaxies

●
Variability increases (larger amplitude and more rapid) as one moves to higher
frequencies.
●
Rapid X-ray Variability on Timescales Down to Minutes!!
●
X-ray variability often implies an emission-region size of light hours or less.
 Emission-Line Variability
C IV Variability in NGC Hβ Variability of Markarian 335
5548

●
The broad emission lines also vary, generally following the continuum with a lag.
●
Leads to the idea of that there’s a physical time delay → different regions between
continuum and EL regions (reverberation mapping)
 Particle Jets

●
A significant minority of
AGNs (about 10%) emit
powerful particle jets.
●
These produce strong radio
emission via synchrotron –
such AGNs are “radio loud”.
Centaurus A
How do we know the jets
are streams of gas?
3C 449

They produce of kind of “skywriting” when

interacting with background

NGC 1265
a “radio trail”
Particle Jets

●
Zooming in on the jet of M87.

●
Note the pointing stability
over a very long timescale.

●
Implies some “gyroscope”
keeping the pointing fixed.

●
Can trace the jet down to the
vicinity of the SMBH.
Courtesy of Monaco
 Super-Luminal Motions
Suppose source emits flashes 1 day apart, while
moving toward you at 0.8c

Flash 1
0.2 lt-day
Flash 1
Flash 2

1 lt-day
Flash 2
0.8c
Flashes emitted 1 day
apart, received
0.8 lt-day
Flash 1 0.2 days apart.

Day 0 Day 1
 Super-Luminal Motions
●
Consider continuously glowing blob, moving
almost directly toward you at 0.8c

0.8 sinΘ
Actual dist. traveled: 0.8 lt-day
0.8 cosΘ
Apparent travel time: (1- 0.8 cosΘ) day
Sideways dist: 0.8 sinΘ
Apparent sideways speed:
0.8c Θ 0.8 sinΘ / (1- 0.8 cosΘ) c

0.8 lt-day

Day 0 Day 1
 Radiative Mechanism
needs to explain:
●
Broad range of luminosities, reaching very large
values.
●
Strong and broad optical/UV emission lines.
●
Emission over a very broad band.
●
Variability.
●
Relativistic Particle jets.
 Radiative Mechanism
needs to explain:
●
Broad range of luminosities, reaching very large
values.
●
Strong and broad optical/UV emission lines.
●
Emission over a very broad band.
●
Variability.
●
Relativistic Particle jets.

Black Hole + Accretion Disk Model

 Sagittarius - A

●
Dynamical mass of Sagittarius – A → few 106 Msun within ~10s AU
 Star proper motions and orbits
Breakthrough from the detection of accelerations (i.e. curvature of proper
motions), and the determination of stellar orbits.
 Dynamical Masses: S2 star
Using Kepler’s Third Law ...

Fitting orbit and radial velocities:

Period ~15.8 y
Eccentricity ~ 0.87
Semi-major axis ~1025 AU
Pericenter ~ 125 AU

MBH ~ 4.3×106 M ⊙

Gillessen +2009
Orbit is not closed for possible proper motion
of the BH. Grey crosses are IR flares

Marconi
 Other Methods
Which are the scales probed in terms of the
Schwarzschild radius, i.e. the BH typical size?

Ferrarese & Ford 2005

 Eddington Limit

●
Accretion produces radiation: radiation makes pressure – can this inhibit further accretion?
●
Radiation pressure acts on electrons; but electrons and ions (protons) cannot separate
because of Coulomb force.

= →L≤
Black Hole + Accretion Disk Model
 Origin of the X-ray Emission

~ Light-minutes scale

●
X-rays are not naturally produced by AGN disks; the disk is too cool.
●
Need to add an accretion disk “corona” with a temperature of ~ 150 keV.
●
This makes X-rays by Compton scattering.
●
Perhaps low-energy X-rays from disk component (soft X-ray excess)
 Obscuration and Radiation
Reprocessing

●
Many active galaxies have obscuring/reprocessing material, often envisioned to be in the form of a “torus”.
●
This material likely produces much of the infrared emission as reprocessed “waste heat” from dust.
 Explaining Emission
Over a Very Broad Band
 Explaining Variability

●
The accretion disk is expected to have about the correct size to explain the
observed variability timescales.
●
It naturally explains the change in variability as a function of frequency.
●
At a fundamental level, the physical origin of the variations is STILL
POORLY UNDERSTOOD!!
●
MHD simulations of accretion disks indicate several possible causes of
variability: local random variations in dissipation, non-axisymmetric
structures, global precession of tilted flows, etc.
●
A deeper understanding will require proper simulation of both dynamics
and thermodynamics.
●
Also can have variable accretion rates, variable obscuration, microlensing.
 Explaining Particle Jets

BH Accretion is a natural good candidate:

●
Relativistic Motions
●
Magnetic Fields (also responsible for the X-ray Corona emission)
●
Stable “gyroscopes”
 Outline
●
Early history
●
AGN basics
●
Finding AGNs
●
AGN terminology
●
AGN unification
●
Dissecting AGNs
●
Focused lecture: Proving the Evidence of a
SuperMassive BH
 Finding AGNs
●
There are many methods for finding AGNs.
●
All methods devised for finding AGNs have
limitations and selection effects.
●
Though some are more effective and give purer
samples than others.
●
For a complete census, want to apply as many
methods as possible enabling cross-checks.
●
Need for Multiple Methods.
 Optical/UV Colors
Stellar
Locus

●
Look for Point Sources brighter in UV than normal stars – Works to z ~ 2-
2.5
At higher redshifts, absorption by the IGM makes Quasars very Red in
Schneider et al. (2010)
●

Blue part of Spectrum

●
These methods work best for unobscured quasars; e.g., reddening
causes trouble.
●
Furthermore, at lower AGN luminosities, host-galaxy light becomes
problematic
BPT Diagram
 Luminosity-Redshift Coverage
from Optical/UV Surveys

Just et al. (2007)

●
The SDSS has now identified ~ 300,000 AGNs.
●
These span a broad range of luminosity and redshift.
●
Note the inevitable luminosity-redshift correlation, common to surveys of all types.
 X-ray Surveys for AGNs
●
X-ray emission nearly universal from AGNs. Almosto all AGNs are STRONG X-ray emission.
●
X-rays have reduced absorption bias compared to optical/UV, especially for high-energy X-rays.
●
X-rays maximize contrast between black-hole vs. host-galaxy light.
●
Can find obscured AGNs and lower luminosity AGNs than in optical/UV.
●
Now are a wide variety of X-ray surveys, ranging from shallow all-sky to deep pencil-beam.

Xue et al. (2011)

 Gamma-Ray Surveys for AGNs

Gamma-ray surveys mostly find AGNs with radio jets pointed right at us,
commonly called “blazars”.
 Infrared Selection of AGNs
●
Several methods have been developed
for the effective selection of AGNs in
sensitive infrared data.
●
Often are seeing infrared power-law
emission or “waste heat” from the AGN
re-emitted by warm dust.
●
These are also relatively resistant
against obscuration effects.
●
They sometimes even find AGNs
missed in X-ray surveys.
●
At lower AGN luminosities, such
surveys suffer substantial
contamination from star-forming
galaxies.
Stern et al. (2012)
 Radio Surveys for AGNs
●
Many of the first quasars were
found via radio selection (3C).
●
About 10% of AGNs are radio-loud
sources.
●
Sensitive radio surveys can detect
many radio-quiet AGNs too.
●
Stars are usually very weak radio
sources, so little stellar
contamination.
●
Generally good radio positions
allow efficient follow-up studies.
●
Often quite incomplete, owing to
radio-quietness of many AGNs.
●
Often involves “human
intervention” → Visual inspection

Typical image from the FIRST survey

 Radio Surveys for AGNs
●
Many of the first quasars were
found via radio selection (3C).
●
About 10% of AGNs are radio-loud
sources.
●
Sensitive radio surveys can detect
many radio-quiet AGNs too.
●
Stars are usually very weak radio
sources, so little stellar
contamination.
●
Generally good radio positions
allow efficient follow-up studies.
●
Often quite incomplete, owing to
radio-quietness of many AGNs.
●
Often involves “human
intervention” → Visual inspection

Typical image from the FIRST survey

The AGN bestiary
G. Risaliti
G. Risaliti
G. Risaliti
Courtesy of Monaco
G. Risaliti
The AGN bestiary
 The AGN Zoo – Key Classification
●
Strength of Radio Emission
– Radio-Loud Quasar vs. Radio-Quiet Quasar
●
Optical/UV Emission-Line Properties
– Seyfert 1 galaxy vs. Seyfert 2 galaxy
– Type 1 Quasar vs. Type 2 Quasar
– Broad Line vs. Narrow Line Radio Galaxy
– Also intermediate Seyfert types, Narrow-Line, Seyfert 1, BL Lac, Weak-Line Quasar
●
Also Variability and UV Absorption-Line Properties
●
Luminosity is also often used in classifications for largely historical reasons; usually not so
fundamental (e.g., Seyferts are just low-luminosity quasars).

Krolik (1999)
Classification Scheme

Krolik (1999)
 Evidence for Unification
●
Key Observation:
Polarized Spectrum of
NGC 1068 (Seyfert 2
archetype) → Shows
clear Broad Lines!
●
Polarized light enhance
the presence of broad
linees.
●
BL in intensity are 100
times fainter than
Continuum and NL
(impossible to detect)
The Unifying Model
 Dissecting AGNs emission
1) The BH region

Courtesy of Pat Hall

 Dissecting AGNs emission
2) The UV/Optical Continuum Emission Region
 Dissecting AGNs emission
3) The Broad Line Region
 Dissecting AGNs emission
4) The Outflow Region
 The Black Hole Region

●
Inner ~50 RS
 The Black Hole Region

●
EHT
 The Black Hole Region

Schematic AGN X-ray Spectral Energy Distribution

Continuum components
●
Power law
●
Soft X-ray excess
●
Compton reflection
hump
Discrete atomic features
●
Iron Kα Line
●
Other line emission
 The Black Hole Region

Schematic AGN X-ray Spectral Energy Distribution

Continuum components
●
Power law
●
Soft X-ray excess
●
Compton reflection
hump
Discrete atomic features
●
Iron Kα Line
●
Other line emission

Suzaku Spectrum Reeves et al. (2007)

 X-ray Power Law
●
Power law N(E) α E-Γ has a photon index of Γ =
1.7-2.2.
●
The “corona” Compton up-scatters UV/optical
photons from the disk to create the power law.
●
Corona likely heated by magnetic flares.
●
Corona has a temperature of ~ 150 keV, beyond
which an ~ exponential cutoff is observed.
●
150 keV “reasonable” → 511 keV pair production
●
The corona’s properties cannot yet be computed
from first principles, but progress being made.
●
Thus the corona’s nature remains uncertain.
– Sandwiching the disk?
– Base of a jet?
 X-ray signature

●
Almost all
AGNs show
some X-ray
emission
 Soft X-ray Excess

●
Strong soft emission of a ~
blackbody spectral form seen
from some objects below ~ 1.5
keV.
●
Too hot and too variable to be
entirely from standard accretion
disk.
●
Likely a combination of disk
emission at lowest energies plus
a cool Compton-scattered
component and disk reflection.
 Compton Reflection Hump

60 keV Compton
<60 keV

absorbed

●
Broad band hump peaking at 20-40 keV.
●
Produced when X-rays shine upon the accretion disk or other material.
●
At low energies have a competition between Compton scattering and photoelectric absorption.
●
Photoelectric absorption dominates below 10-15~ keV
●
At high energies drops off due to e.g. Compton recoil, and the power-law cutoff.
●
Affected by Doppler shifts, beaming, GR
 Iron Kα Line ~ 2keV ~ c/3

Fabian et al. (2000)

6.4 keV
●
Made via iron fluorescence when disk irradiated by X-rays.
●
Iron has best product of abundance and fluorescent yield.
●
With very high-S/N data, can use to estimate disk
inclination, disk emissivity, and black-hole spin.
 X-ray Variability

●
AGN variability power spectra usually do not show periodicity or quasi-
periodicity.
●
Observe noise power spectra with a bend/break at ~ 0.1-100 days.
●
The bend/break timescale can be correlated with other AGN physical
properties.
 X-ray Variability

McHardy et al. (2005, 2006)

●
AGN variability power spectra usually do not show periodicity or quasi-
periodicity.
●
Observe noise power spectra with a bend/break at ~ 0.1-100 days.
●
The bend/break timescale can be correlated with other AGN physical
properties.
 The Broad Line Region (BLR)
●
Relative strengths of emission lines in AGN spectra indicate we are
observing gas in photoionization equilibrium at ~ 10 4 K.
●
Abundances about solar or slightly supersolar.
●
Lines are kinematically composite:
– Broad components
●
Doppler widths of 1000-25000 km/s.
● Arise in gas with density ne ~109 -1011 cm -3 (as determined from strengths (missing) of certain
density sensitive lines like [O III] and CIII]).
●
We called it the “Broad Line Region”
●
These motions are not thermal (~10 km/s for 10 4 K).
●
Rather are supersonic bulk motions.
●
The larger Doppler widths of the broad lines indicate they arise deeper in
the gravitational potential.
Strong and Broad Lines Common
 Broad Lines Lag Continuum
●
The broad lines vary on short
timescales, usually following the
continuum variations with a time
delay.
●
Thought to be the light travel time
across the BLR, leading to the idea
of “reverberation mapping”.
●
Provides a way to measure the BLR
size, which is ~ 5-500 light days
depending upon luminosity.
●
The narrow lines do not vary on
short timescales, indicating they are
from a much larger region.

Grier et al. (2012)

 Size of the Broad Line Region
Kaspi et al. (2005)

Bentz et al. (2013)

● The BLR radius is found to scale as RBLR ~ L0.5-0.7 .

●
The slope is about as expected if all AGNs have similar ionization
parameters and densities in their BLRs (r ~ L 0.5 )
Derived from → ξ = L/(n R2)
Ionization parameter
 Object-to-Object Differences
in Broad Line Properties
●
There are significant object-to-object differences in broad-line properties.
●
There is a set of emission-line properties that vary in a correlated manner
called “Eigenvector 1”.
● Likely related to L / LEdd .

Eigenvector 1 from PCA

Boroson & Green (1992)

 Reverberation Mapping:
Stratification and Virialization
●
Reverberation lags have now been measured for ~ 50-100 AGNs.
●
The current sample is biased toward AGNs with relatively strong lines.
● Mostly measured for Hβ, but in some cases for multiple lines.
●
The highest ionization emission lines respond most rapidly to continuum
changes, indicating ionization stratification.
●
A plot of line width vs. lag (τ) shows that v ~ τ-0.5 , as expected for virialized
gas dominated by the gravitational potential of the central source.

Other lines
2022

Hβ
 Estimating Black Hole Masses
●
Can estimate black-hole masses
following the virial theorem:

Virial Expectation

●
Where f is a factor that includes
Other lines (unknown) BLR geometry and
inclination.
●
Comparison with other mass-
estimation methods indicates an
average value of f ~ 4-5.
●
Masses measured this way
appear accurate to within a factor
Hβ of ~ 3 when Hβ is used.
●
Note this method can be used, if
patient, for masses at high
redshifts.
 Mass-Luminosity Relationship
Peterson et al. (2004) ● Can combine the R BLR -L relation with the
virial theorem to estimate single-epoch
masses. For example…

●
Similar relations exist for Mg II and C IV.
●
These allow quick estimates for large AGN
samples, but their accuracy is no better
than a factor of several. The main
challenge is characterizing the line widths,
where caution is needed.
●
Statistical use of such masses in large
samples is probably OK, but individual
mass estimates may be unreliable.
 What is the Nature of the BLR?
●
After much research, it is appearing increasingly likely that the
BLR itself has a composite nature:
●
Moderate-ionization and high-optical-depth region
– Largely responsible for the Balmer-line emission and Mg II
– Accretion disk itself?
– A disk with a large line-emitting region can make single-peaked
profiles consistent with most objects
●
High-ionization and moderate-optical-depth region
– Largely responsible for the high-ionization lines
– Accretion-disk wind?
– Helps explain blueshifts of high-ionization lines and blueward line
asymmetries
 Outflowing
Winds
●
Blueshifted UV Broad Absorption Lines (BAL) in Quasars
●
Seen in 10-15 % of sources, but thought to be associated to 20-30% of sources (due to selection
effects)
●
Complex and multiple absorption features
●
Defined to be broader than 2000 km/s
●
Smaller are often defined as mini-BALs

Arav et al. (2001)

 UV Outflows Found Over a
Wide AGN Luminosity Range

●
There’s seem to be a correlation with L:
– The larger the luminosity, the larger the wind velocity
 BAL Fraction Depends
Upon Radio Power
Shankar et al. (2008)
●
BALs generally avoid highly
radio-luminous quasars (though
not entirely).
●
Reason for this is still not
entirely clear.
●
Wind-jet connection?
– Connection with Eddington
Rate (ER)?
– Lower ER → Radio Mode
– Higher ER → Quasar Mode
●
Orientation effects?
 Wind Model

●
Most AGNs likely drive winds – orientation effect.
●
Multiple lines from same transition probe distinct wind components.
●
Wind material exists over a wide range of radii from 0.01 pc to kpc
scales. Often outside the BLR.
●
Lines often saturated with partial covering.
●
Wind is driven by radiative pressure on the lines from the Equatorial
Accretion-Disk or Torus to Velocities of ~ 100-30000 km/s
 Effect of Winds
●
Why Care About AGN Winds?
●
Significantly affect observed AGN properties (UV line
absorption, high-ionization line emission, reddening,
polarization, X-ray absorption).
●
Substantial part of typical AGN nuclear regions; seen in
absorption in ~ 30+% of AGNs.
●
Help black-hole accretion to proceed by removing angular
momentum from the disk.
●
Can evacuate gas from the host galaxy, likely affecting
black-hole growth and galaxy evolution.
Wind Feedback into the ISM
The Narrow Line Region
 Imaging the NLR
●
NGC 1068 (Fath 1908)
NGC 1068 ●
The line emission region is
clumpy and complex.
●
NLR is clearly not spherically
symmetric, but rather is
roughly axisymmetric.
●
NLR axis generally coincides
with radio axis in cases where
extended linear radio emission
is detected.
●
In some sources, we see
strong line emission from
regions where the radio jet is
colliding with the ISM and
causing shocks – an additional
source of ionization.

HST - Macchetto et al. (1994)

 The Narrow Line Region
●
Narrow components
– Doppler widths typically less than 900 km/s.
– Arise in relatively low-density gas (ne ~ 103 cm-3 ) (Forbidden Lines!!)
– From the “Narrow Line Region”.
●
Largest spatial scale where ionizing radiation from the AGN dominates.
●
NLR can be spatially resolved in the optical; has sizes of ~ 100+ pc in local Seyferts (and even larger
in quasars).
●
Can map out physical and kinematic properties directly to some extent.
●
FWHM values are 200-900 km s/1 , with line profiles varying across NLR.
●
See a wide range of ionization states:
– Low ionization (e.g., [O I] λ6300)
– High ionization (e.g., [O III] λ4959, 5007)
– Sometimes even very highly ionized species (e.g., iron coronal lines)

●
From line ratios, infer that the NLR is mostly photoionized by the AGN continuum (with some likely
additional ionization by shocks from radio jets).
●
NLR line EWs drop with increasing continuum luminosity, and are often undetectable in high-luminosity
quasars.
●
NLR becoming larger than the host galaxy?
If inoization parameter ξ is constant → ξ = L/(n R2) → R Can exeed the Galaxy (running out of gas)
 Importance of NLR
●
Line peaks provide useful systemic
redshifts for AGNs.
●
Useful as a bolometer for inferring
AGN total power.
– NLR lines can be used to estimate
rough bolometric luminosities, even for
obscured AGNs.
– Emitted from a region larger than any
nuclear obscuration → Vary on longer
timescales (decades).
●
Connection with Eigenvector 1.
●
NLR line widths are correlated with
host-galaxy bulge luminosity (and
bulge gravitational potential).
●
Anisotropic illumination provides
clues about AGN geometry and
orientation.
 Importance of NLR
●
Line peaks provide useful systemic
redshifts for AGNs.
●
Useful as a bolometer for inferring
AGN total power.
– NLR lines can be used to estimate
rough bolometric luminosities, even
for obscured AGNs.
– Emitted from a region larger than any
nuclear obscuration.
●
Connection with Eigenvector 1.
●
NLR line widths are correlated with
host-galaxy bulge luminosity (and
bulge gravitational potential).
●
Anisotropic illumination provides Marconi et al. (1994)
clues about AGN geometry and
orientation.
 Importance of NLR
●
Line peaks provide useful systemic
redshifts for AGNs.
●
Useful as a bolometer for inferring
AGN total power.
– NLR lines can be used to estimate
rough bolometric luminosities, even
for obscured AGNs.
– Emitted from a region larger than any
nuclear obscuration.
●
Connection with Eigenvector 1.
●
NLR line widths are correlated with
host-galaxy bulge luminosity (and
bulge gravitational potential). Netzer (2013)
●
Anisotropic illumination provides
Orientation of the Cone not correlated
clues about AGN geometry and with host galaxy
orientation.
The Torus and the Unifying Model
●
To first approximation, the optical/UV
spectra of AGNs separate into two broad
spectral types:
●
Type 1:
– Broad permitted emission lines, particularly the
Balmer lines
– Permitted lines clearly broader than forbidden
lines
– Moderate EW forbidden lines
●
Type 2:
– Narrow permitted emission lines, particularly
the Balmer lines
– Permitted lines have similar widths to forbidden
lines
– High EW forbidden lines
●
There are additional complications to this
simple scheme (e.g., intermediate type
classifications, narrow-line Type 1 AGNs).
●
These optical spectral differences have
come to be understood as (often) due to
orientation-dependent central obscuration
by a so-called “TORUS”.
The Torus and the Unifying Model
●
The torus is presumed to be an axisymmetric structure of
large height so that, at least at low luminosities, the
majority of AGNs are obscured by it.
●
It is made of a combination of dusty atomic and
molecular gas.
●
The dust causes large extinction in the optical/UV and
sometimes even in the NIR.
●
The torus lies between the BLR and the NLR.
●
Type 2 AGNs are those obscured by the torus, and
emission on the scale of the BLR and smaller is
obscured by it.
●
Models explaining differences between Type 1 and Type
2 AGNs this way are referred to as “unification models”.
●
These models have had much success, though they are
not complete and there are likely exceptions.
●
There appear to be substantial object-to-object variations
in the covering factor and geometry of the torus.
●
Keplerian velocities at the torus distance are ~ 1000 km/s
●
Density of torus “clumps” are ~ 105 -107 cm-3 .
●
The estimated mass of the torus is only a small fraction
of the SMBH mass.
●
The torus is likely a dynamic system (maybe clumpy)
being part of the general flow of matter from the galaxy’s
center to the SMBH. But the details remain unclear.
 Size of the Torus

Burtscher et al. (2013)

●
We know the torus must lie between the BLR and NLR, but we can be more specific.
●
We can now directly measure the size of the torus using
– Dust reverberation mapping between the V-band and K-band light curves.
– Interferometry in the NIR and MIR mapping the hot inner wall and the cool dust.
●
The size of the torus appears to scale as L0.5 (consistent with constant ionization parameter).
●
The inner edge of the torus is at about 3 times the BLR radius for Hβ as determined from reverberation mapping.
 Jets

Could be
several
Mpc (this
one 3 Mpc)

Palma et al. (2000)

Jets properties
●
Jets often appear relativistic based
on beaming, apparent superluminal
motions on sub-pc scales, and
variability.
●
Thus, the apparent properties of a
jet depend strongly upon orientation.
●
At least in some cases, jets are
launched on very small scales
(making much of the “core” radio
emission). And they can be
collimated over a huge range of
scales.
●
Jets are often seen in the X-ray and
also the optical.
 Flat vs. Steep Spectrum
Radio Sources
● Often observe radio power-law spectra: L ν ~ ν-α
●
Radio-loud AGNs with dominant radio cores usually show “flat” radio
spectra with α < 0.5
●
Radio-loud AGNs with dominant lobes usually show “steep” radio spectra
with α> 0.5
●
Much of the difference in measured value of α is due to inclination of
radio jet to our line-of-sight

Gupta et al. 2017

 How Are Jets Made?
●
Generally invoke MHD processes to
divert some of the inflowing plasma
outward and then keep it collimated.
●
But exactly how to combine the
“ingredients” remains poorly understood.
●
What sets if a strong jet will be
launched?
– SMBH spin?
– Magnetic geometry?
– Environment?
– Connection with Eddington Rate (ER)?
– Lower ER → Radio Mode
– Higher ER → Quasar Mode
 Jet Feedback in Clusters

McNamara et al. (2005)

Fabian et al. (2003)

Jets can do substantial work against the hot gas in galaxy clusters
AGN Feedback in Clusters

Fabjan+10